Dear Steve, Paco et al.
> IMHO the above paragraph cannot really be said to be a requirement :-\
> . It begins with a proposed implementation ("may have a dimension
> which serves as an index") and follows up with some of the
> functionality that can (and cannot) easily be achieved through that
> implementation.
Yes, I agree, I should have taken a step back from the index dimension. So
I would rephrase the requirement as follows:
A means is needed (a) to aggregate data variables from the members of an
ensemble, derived from different models, integrations, institutions supplying
the data, etc., where the data from each ensemble member is a function of the
same spatiotemporal coordinates and/or other physical independent variables;
(b) to provide metadata identifying the ensemble members.
Then the proposed implementation begins:
We propose to introduce a dimension for the data variable to serve as an
index over the ensemble members.
and continues by proposing auxiliary coordinate variables, etc., as in my
last posting
http://www.cgd.ucar.edu/pipermail/cf-metadata/2006/001397.html
The requirement covers the areas you (Steve) have given (in more detail) in
your points 1-3. Your requirements 4-6, however, are about functionality which
programs processing the data might need to have. While I agree we should bear
in mind the convenience of processing the proposed metadata, and hence that is
a consideration in deciding which implementation to choose, I don't think it
forms part of the requirement. CF is a metadata convention, not a specification
for data processing software.
Paco, does the present proposal meet your requirements? If so, I would urge
that we adopt it. We still have the question about whether the new metadata
items should be given standard_names, or some other atttribute. I would suggest
that they should be standard_names unless anyone can propose both a requirement
for a distinction and a reliable objective test which defines that distinction.
Best wishes
Jonathan
Received on Sat Dec 23 2006 - 14:03:47 GMT
